pipeline-scorer

Score PR review findings with a quantitative rubric and output routing decisions.

2|1|Updated Dec 25, 2025

Search Tags:#automation #scoring #rubric #triage #code-review #pr-review

Ready to run

npx skills add https://github.com/solo-ist/prose --skill pipeline-scorer

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: pipeline-scorer
Source: https://github.com/solo-ist/prose/tree/main/.claude/skills/pipeline-scorer
Command: npx skills add https://github.com/solo-ist/prose --skill pipeline-scorer

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Score PR review findings using a quantitative rubric to inform triage and routing decisions for software changes.

Core Features & Use Cases

Six-dimension scoring delivering a 1-10 composite score.
Automated routing guidance: auto-fix, review, or complex review based on threshold.
Output includes per-dimension scores and a machine-readable sentinel for downstream pipelines.

Quick Start

Run the pipeline-scorer on a PR number to obtain a structured evaluation and routing guidance.

Frequently Asked Questions about pipeline-scorer

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I automate PR triage with a quantitative scoring rubric?▼

Automate PR triage by applying a six-dimension quantitative rubric to review findings, generating a 1-10 composite score. This provides machine-readable routing decisions to quickly classify pull requests across varying change types and risk profiles.

What is the best way to route code review findings for large pull requests?▼

Routing code review findings for large PRs is handled by evaluating a composite 1-10 score against defined thresholds. The scorer outputs automated guidance to categorize each PR for auto-fix, standard review, or complex review.

Can I use pipeline-scorer for machine-readable code review routing decisions?▼

Yes, pipeline-scorer generates a machine-readable sentinel containing per-dimension scores and a routing decision. This output is designed specifically for consumption by downstream automation pipelines.

Does automated PR scoring work for both small and large change risk profiles?▼

Automated PR scoring applies to both small and large pull requests with varying change types and risk profiles. The quantitative rubric adjusts to the specific findings to ensure reliable routing regardless of PR size.

How do PR review scores determine if a pull request needs a complex review?▼

PR review scores determine review complexity by comparing the 1-10 composite score against a threshold. Scores that fall below the auto-fix or standard review thresholds trigger a complex review routing decision.

What limitations exist when triaging code review findings with a quantitative rubric?▼

Triage with a quantitative rubric relies solely on identified PR review findings to generate a 1-10 score. It does not account for context outside the provided findings when outputting routing decisions for downstream pipelines.